Factors controlling the net ecosystem production of cryoconite on Western Himalayan glaciers

In situ experiments were conducted to determine the net ecosystem production (NEP) in cryoconite holes from the surface of two glaciers (Patsio glacier and Chhota Shigri glacier) in the Western Himalaya during the melt season from August to September 2019. The study aimed to gain an insight into the factors controlling microbial activity on glacier surfaces in this region. A wide range of parameters, including sediment thickness, TOC %, TN %, chlorophyll-a concentration, altitudinal position, and grain size of the cryoconite mineral particles were considered as potential controlling factors. From redundancy analysis, the rate of Respiration observed in cryoconite at Chhota Shigri glacier was predominantly explained by sediment thickness in cryoconite holes (37.1% of the total variance, p < 0.05) with Photosynthesis largely explained by the chlorophyll-a content of the sediment (39.6%, p < 0.05). NEP was explained primarily by the TOC content and sediment thickness in cryoconite holes (35.8% and 22.1% respectively, p < 0.05). The altitudinal position of the cryoconite is strongly correlated with biological activity, suggesting that the stability of cryoconite holes was an important factor driving primary productivity and respiration rate on the surface of Chhota Shigri glacier. We calculated that the number of melt seasons required to accumulate organic carbon in thin sediment layers (< 0.3 cm), based on our measured NEP rates, ranged from 11 to 70 years, indicating that the organic carbon in cryoconite holes largely derives from allochthonous inputs, such as elsewhere on the glacier surface. Phototrophic biomass in the same thin sediment layer of cryoconite was estimated to take atleast 4 months to be produced in situ (with mean estimated time upto 1.7 ± 1.5 years). Organic matter accumulated inside the cryoconite holes both through allochthonous deposition and via biological activity on the glacier surface in these areas may have the potential to export dissolved organic matter and associated nutrients to downstream ecosystems. Given the importance of Himalayan glaciers as a vital water source for millions of people downstream, this study highlights the need for further investigation in aspects of the quantification of in situ produced organic matter and its impact on supraglacial melting in the Himalaya.

considered as potential controlling factors. From redundancy analysis, the rate of Respiration observed in cryoconite at Chhota Shigri glacier was predominantly explained by sediment thickness in cryoconite holes (37.1% of the total variance, p < 0.05) with Photosynthesis largely explained by the chlorophyll-a content of the sediment (39.6%, p < 0.05). NEP was explained primarily by the TOC content and sediment thickness in cryoconite holes (35.8% and 22.1% respectively, p < 0.05). The altitudinal position of the cryoconite is strongly correlated with biological activity, suggesting that the stability of cryoconite holes was an important factor driving primary productivity and respiration rate on the surface of Chhota Shigri glacier. We calculated that the number of melt seasons required to accumulate organic carbon in thin sediment layers (< 0.3 cm), based on our measured NEP rates, ranged from 11 to 70 years, indicating that the organic carbon in cryoconite holes largely derives from allochthonous inputs, such as elsewhere on the glacier surface. Phototrophic biomass in the same thin sediment layer of cryoconite was estimated to take atleast 4 months to be produced in situ (with mean estimated time upto 1.7 ± 1.5 years). Organic matter accumulated inside the cryoconite holes both through allochthonous deposition and via biological activity on the glacier surface in these areas may have the potential to export dissolved organic matter and associated nutrients to downstream ecosystems. Given the importance of Himalayan glaciers as a vital water source for millions of people downstream, this

Introduction
The Himalayan region has experienced a 40% loss of its glacier area since the Little Ice Age maximum (1300 CE to 1600 CE), and glaciers melting has accelerated by 10 times in the past few decades (Lee et al. 2021). Several studies have reported the accelerating loss of ice across Himalayan glaciers due to a changing climate (Fujita and Nuimura 2011;Bolch et al. 2012;Azam et al. 2014a;Zhao et al. 2016;Mandal et al. 2016;Maurer et al. 2019;Garg et al. 2021). Future projections for high-altitude Asian glaciers indicate at least a 50% loss of current ice mass and volume by mid-century as climate change intensifies (Zhao et al. 2014;Rounce et al. 2020). The loss of glacier area and volume across the Himalayan region is of particular concern, since it has both social and environmental implications, with a number of challenges associated with water supplies and future aggravated ecosystem and environmental degradation (Bakke et al. 2016;Pritchard 2017;Higgins et al. 2018;Bolch et al. 2019).
The supraglacial environment is the uppermost part on the glacier surface that interacts directly with the atmosphere and receives direct solar radiation and thus responds to external climate forces through accelerating surface melting. One of the potential factors which might accelerate the rapid decline of glacier ice areas and volume due to atmospheric warming and precipitation changes is a reduction of surface albedo due to the accumulation of organic and inorganic particles called "cryoconite" in the supraglacial environment (Takeuchi et al. , 2001(Takeuchi et al. , 2010Hodson et al. 2010b). Cryoconite is produced locally from the melt-out of glacial sediments, inputs of aeolian dust, alongside biological productivity and can be found disperse on snow and ice surfaces, thereby causing the darkening of glaciers (Takeuchi et al. 2001;Remias et al. 2005;Qian et al. 2015;Musilova et al. 2016;Stibal et al. 2017). Thin cryoconite layers on glacier surfaces has the potential to absorb solar radiation due to its low reflectivity, and can form cylindrical depressions on the glacier surface known as 'cryoconite holes' Telling et al. 2012). These serve as habitats for autotrophic microorganisms, including bacteria, algae (Anesio et al. 2009) and as well as other heterotrophic microbial communities in the wider glacial ecosystem (Zarsky et al. 2013;Nicholes et al. 2019;Rozwalak et al. 2022). Due to the presence of liquid water, light, and nutrients (Stibal et al. 2008) during the ablation period, microorganisms in cryoconite holes can display significant rates of metabolism (Foreman et al. 2007;Edwards et al. 2011) with high primary production and respiration rates that are comparable to those of eutrophic ecosystems in warmer regions (Anesio et al. 2009). Cryoconite holes also provide a protective aquatic environment on glacier surfaces for microorganisms from extreme UV radiation, temperature fluctuations and changes in environmental conditions by forming ice lids over the granules (Hodson et al. 2010b;Langford et al. 2014;Bagshaw et al. 2016). However, depending on the time of the year, surface temperature variability and meltwater connectivity, the structure and microbial abundance within the cryoconite holes may vary both temporally and spatially (Lutz et al. 2017;Pittino et al. 2018), which may affect the microbial activity on the glacier surface. Microorganisms trapped in cryoconite holes often display high net biological productivity, resulting in organic matter accumulation in supraglacial ecosystems (Hodson et al. 2007;Telling et al. 2012). This produces dark-coloured material which can further reduce the albedo of the glacial surface, a process known as the 'bio-albedo effect' (Takeuchi et al. 2001;Cook et al. 2017). However, studies investigating these biological-melt feedbacks have largely been conducted on Arctic and Antarctic glaciers (Säwström et al. 2002;Hodson et al. 2005;Stibal et al. 2008;Uetake et al. 2010;MacDonell and Fitzsimons 2012;Zarsky et al. 2013;Bagshaw, et al. 2013;Lutz et al. 2017;Holland et al. 2019;Leidman et al. 2021).
Previous studies in the Himalaya have demonstrated the presence of diverse microbial communities on glacier surfaces (Sanyal et al. 2018), and inferred the biological production of organic matter via biomarkers and geochemical tracers (Nizam et al. 2020). Organic matter formed on the glacier surface can be largely autochthonous produced through in situ photosynthesis on the glacier surface (Anesio et al. 2009;Telling et al. 2010Telling et al. , 2012. A similar conclusion was drawn by Nizam et al. (2020) in their study on glacier surfaces of the Western Himalaya, however, their study lacked experimental data to confirm the operation of biological processes, or their potential direct impact on organic matter accumulation, and potential surface darkening. Here we conducted a suite of experiments to directly determine for the first time the potential biological productivity of microorganisms in cryoconite holes of two glaciers, Patsio and Chhota Shigri glaciers in the Western Indian Himalaya, with an emphasis on identifying the factors controlling the net ecosystem production (NEP) in surface microbial ecosystems.

Measurements of in situ net ecosystem production and respiration
Field experiments were performed in August-September 2019 to determine rates of Respiration and Net Ecosystem Production in cryoconite holes on both the surfaces of Chhota Shigri and Patsio Glaciers. A total of 16 sites (cryoconite holes) were chosen on both glaciers for in situ experiments (12 cryoconite sites in Chhota Shigri and 4 cryoconite sites in Patsio were intended; however, two of the experiments were lost due to breakages). The experiments were performed over a large altitudinal gradient in Chhota Shigri glacier's ablation zone (between ~ 4500 and 4800 m a.s.l.) and at 5050 m a.s.l. in Patsio glacier. Cleaned and autoclaved BOD bottles of 30 mL were used for the experiments. One BOD bottle was covered with aluminium foil to represent (dark only) Respiration, while another was left uncovered to represent NEP (light only). Both bottles were filled with approximately the same amount of sediment, with a thickness range of 0.2-0.8 cm measured using a measuring tape to represent typical in situ thicknesses at the sampling sites. Similarly, one dark and one light blank bottle filled with stream water containing no sediment were taken to observe the influence of water only during Respiration and NEP respectively. Using a portable DO meter (Hach HQ40d DO meter), the initial concentration of dissolved oxygen (mg L − 1 ) was measured in all bottles. The BOD bottles were then left on the glacier surface completely submersed inside the cryoconite holes water. Where the holes were broken, the bottles were left exposed on the glacier surface in bare cryoconite. Final dissolved oxygen concentrations were measured after 24 ± 2 h. Using the light and dark incubation, the rates of NEP and Respiration respectively were directly estimated (Hodson et al. 2010a;Telling et al. 2012) in units of µg C L − 1 day − 1 . Measurements were corrected with the dry sediment weight of cryoconite and converted to µg C g − 1 day − 1 . The primary production or Photosynthesis rate (autotrophy) is calculated as the sum of NEP and Respiration (Hodson et al. 2010a) where P = rate of Photosynthesis, R = rate of Respiration in µg C L − 1 day − 1 (or µg C g − 1 day − 1 ). (1) Chlorophyll-a content in the cryoconite sediments Cryoconite samples were stored soon after collection in a portable field freezer at sub-zero (− 10 to − 20 °C) temperatures in the dark and brought to the LOWTEX laboratory, University of Bristol. They were stored frozen until analysis which was conducted within 3 to 4 months. All extraction procedures for chlorophyll-a were conducted in low light conditions to avoid degradation. Frozen cryoconite sediments were weighed in pre-weighed 15 mL Falcon tubes covered with aluminium foil. The initial weight of the sediment was noted to determine the moisture content, and in order to later normalise the pigment concentration to the dry weight of sediment. Chlorophyll extraction was performed using 90% acetone which was added to the pre-weighed Falcon tubes to completely cover the sediment, and the solution was mixed for two minutes on a vortex shaker. The Falcon tubes were then sonicated in an icechilled sonicator bath (> 10 °C) for 20 min and kept at − 4 °C for 24 h. The tubes were vortexed again for 5 min, followed by centrifugation at 1500 rpm. The supernatants were then filtered through a 0.2 μm Whatman Puradisc Nylon membrane syringe filters and subsequently analysed for the chlorophyll-a concentration using a UV-vis spectrophotometer. To report chlorophyll-a, as corrected for pheophytin content, absorption both before and after acidification of the extracts was measured at wavelength 665 nm and another at 750 nm for turbidity correction. Estimation of the chlorophyll-a concentration was conducted following Lorenzen (1967) and normalised to the dry weight of sediment. The limit of detection of the chlorophylla calculated using six blank acetone samples was 0.03 ± 0.17 µg L − 1 .
Grain size distribution of the cryoconite minerals Dry sediment samples were combusted in a furnace for 3 h at 950 °C to remove organic carbon and carbonated matter. Particles larger than 2 mm were removed with a sieve after loosely separating the particles with a pestle and mortar. The sample grain sizes were then measured using a Malvern MS3000 Mastersizer. Analyses were conducted in triplicate for each sample and a blank sample (water only) was run in between sample analyses. The mean value of the triplicate sample was reported for the grain size distribution.

Elemental composition of the cryoconite of Chhota Shigri
The carbon, nitrogen and sulfur content of cryoconite from all sites were analyzed using a CHNS Vario PYRO cube Elemental Analyzer. The sediment was dried at 108 °C overnight, finely ground with a pestle and mortar, and passed through a 150 mesh sieve. Approximately 10 mg of the untreated samples were carefully sealed in aluminium capsules to measure the total carbon (TC) content. The total organic carbon (TOC) content of the cryoconite constituted averaged 98% of the TC in a sub-set of cryoconite sediment  (Table 2). Due to the negligible total inorganic carbon (TIC) content, TC for the present study were reported as TOC.
The instrumental limit of detection was 0.001% or 10 ppm.

Statistical analysis
Multivariate statistical analyses were performed using the software package R (version 4.0.0). Redundancy discriminant analysis (RDA) was used to determine the impact of independent variables (explanatory variables) on dependent variables (response variables) in the data set. This statistical methodology extracts and summarizes the variation of the response variable that can be explained by a series of explanatory variables (Borcard et al. 2011).

Physical and chemical properties of cryoconite holes
Cryoconite hole diameter, depth, water depth and sediment thickness were measured at the experimental sites at Chhota Shigri and Patsio glaciers ( Table 1). The sampling sites for in situ cryoconite in Chhota Shigri glacier were grouped according to elevation as "lower", "middle" and 'upper' sites, which corresponded to glacier slopes of 13˚ − 23˚, 10˚ − 19˚ and 10˚ − 12˚ respectively. Cryoconite holes observed in the lower site in Chhota Shigri (4562 m a.s.l.) were observed to be mostly broken, and scattered due to the connection of multiple channels and sediment exposed on the ice surface ( Fig. 2a1, a2). Cryoconite holes at middle elevation sites of Chhota Shigri (4678-4747 m a.s.l.) were larger in size and possibly formed due to fusion of several cryoconite holes to form a larger hole (Fig. 2b) . Multiple surface streams were also observed alongside at this site. At the upper sites in the Chhota Shigri and Patsio Glaciers, the cryoconite holes were circular shaped and smaller in size compared to the middle sites and were relatively stable and hydrologically isolated (Fig. 2c,d).  A summary of the parameters measured is given in Table 3. A scatter matrix diagram comprising a scatter plot, distribution diagram, boxplot, and Pearson correlation of all the parameters is given in Fig. 3.
The TOC and TN content of cryoconite showed an overall significant positive correlation with chlorophyll-a (df = 10, r = 0.94 and 0.94 respectively, p < 0.01 (Fig. 3), R 2 = 0.88 and 0.89 respectively, p < 0.001 (Additional File Fig. 2A2)). The TOC, TN and chlorophyll-a content of the cryoconite increased from the lower and middle to the upper sites in the Chhota Shigri Glacier (Fig. 3), with mean values increasing from 0.52 to 0.68 to 1.81% for TOC, 0.06 and 0.08 to 0.20% for TN and 3.62 and 5.3 to 8.4 µg g − 1 for chlorophyll-a content from lower and middle to upper sites (Table 3). In Patsio glacier, the observed mean TOC (2.38%), TN (0.21%) and chlorophyll-a content (12.6 µg g − 1 ) were close to that of the upper sites at Chhota Shigri Glacier. In addition, the mean grain size of the cryoconite minerals in the upper region at Chhota Shigri and at Patsio glaciers (39.60 and 45.65 μm respectively) was smaller than that of the lower (62.88 μm) and middle (63.30 μm) sites at the Chhota Shigri glacier. The TOC content of cryoconite lacked a statistically significant association with sediment thickness (r = 0.16, df = 10, p > 0.05); but did show a significant negative correlation with the mean grain size of the cryoconite mineral particulate (TOC, R 2 = 0.59, slope = − 23, df = 10, p = 0.002 (Additional File)). Similarly, TN and chlorophyll-a were significantly negatively associated with the mean grain size of cryoconite mineral particulate matter (TN and chlorophyll-a, R 2 = 0.61, p = 0.0017 and 0.63, p = 0.0013 respectively, df = 10 (Additional File)). The mean C:N ratios in the cryoconite were 8.9-11.3, which is similar to well-decomposed organic matter as observed in glaciers in the Himalayas and Arctic (Takeuchi 2002). The mean S:C ratios were 0.01-0.07, largely consistent with ratios reported for microbial cells (range 0.016 to 0.084, Fagerbakke et al., 1996).

Measurement of microbial productivity
The rate of Photosynthesis calculated from the 24 ± 2 h incubations in the BOD bottles containing cryoconite at the lower, middle and upper experimental sites at Chhota Shigri and Patsio glacier ranged from 180 to 375 µg C L − 1 day − 1 , 210 to 491 µg C L − 1 day − 1 , 588 to 937 µg C L − 1 day − 1 and 1162.5 to 1200 µg C L − 1 day − 1 respectively, with mean values given in Table 4. The rate of Respiration ranged from 225 to 405 µg C L − 1 day − 1 (lower sites), 378 to 656 µg C L − 1 day − 1 (middle sites), 401 to 825 µg C L − 1 day − 1 (upper sites) at Chhota Shigri and 521 to 577 µg C L − 1 day − 1 at Patsio glacier. The rates of NEP were − 217 to 123 µg C L − 1 day − 1 (lower sites), -206 to -22 µg C L − 1 day − 1 (middle sites) and 112 to 236 µg C L − 1 day − 1 (upper sites) and 585 to 678 µg C L − 1 day − 1 (Patsio). On average, rates of Photosynthesis and Respiration in the water only incubations were 3 to 4 times lower than in those in the bottles containing cryoconite (Table 4). In general, sediment layers of thickness of approximately < 0.3 cm were generally dominated by net autotrophy (mean of 251 µg C L − 1 day − 1 ) while cryoconite with thicknesses of ≥ 0.5 cm were dominated by net heterotrophy (mean of -163 µg C L − 1 day − 1 ) ( Fig. 4) at both glaciers. Cryoconite at the upper sites and Patsio glacier displayed higher rates of Photosynthesis and Respiration compared to the lower and middle sites at Chhota Shigri (Table 4).

Multivariate statistical analysis
To infer the controls upon microbial activity in the cryoconite holes, multivariate statistical analyses were performed. NEP, Photosynthesis and Respiration were taken as dependent variables with sediment thickness, TOC %, TN %, Chlorophyll-a, elevation, and grain size as independent variables. Boxplots of Respiration, NEP and Photosynthesis are presented in Fig. 5. Detrended correspondence analysis (DCA) was performed in order to determine the appropriate model to describe the association of the variables. The length of the axis ordination was 0.85 SD, therefore, redundancy analysis was determined as a suitable model to define the relationship between the variables. Before performing the ordination, all variables were standardized prior to the RDA. Since TN showed high collinearity with other parameters, it was therefore excluded from the model calculation to prevent overfitting of the model. The result of the RDA plot is given in Fig. 6 with manual forward selection using NEP, Photosynthesis and Respiration as dependent variables. Based on the multivariate analysis, sediment thickness, altitude, chlorophylla and TOC were identified as the highly significant Fig. 4 Association between respiration and photosynthesis in cryoconite and water only (stream water without cryoconite) at different elevation sites at CS Chhota Shigri and Patsio glaciers   (Table 5).

Factors controlling the microbial activity on the Himalayan glacial surface
The results of our study are consistent with previous work in other glaciated regions which has demonstrated that altitude, sediment grain size, cryoconite hole stability and other physical characteristics of the cryoconite on the glacier surfaces are important influencers of biological productivity in cryoconite holes (Langford et al. 2014;Chandler et al. 2015). In the upper sites at Chhota Shigri and Patsio glaciers, the cryoconite holes were hydrologically isolated, therefore the limited water movement and relative stability of the cryoconite in these sites may favour microbial growth (Langford et al. 2014). This was supported by the observed higher concentration of TOC, TN and chlorophyll-a content in these sites compare to lower sites where cryoconite holes were generally exposed and washed out due to greater hydrological connectivity. Also significant association of grain size of the cryoconite with TOC (R 2 = 0.59, df = 10, p = 0.002), TN (R 2 = 0.61, df = 10, p = 0.0017) and chlorophylla (R 2 = 0.63, df = 10, p = 0.0013) (Additional Files Fig. 3) was observed in our study. Fine cryoconite mineral particles dominated in the upper sites in Chhota Shigri and Patsio and the concentrations of Fig. 6 Result of RDA plot of the explanatory variables (sediment thickness, elevation, chlorophyll-a and TOC) with Respiration, Photosynthesis and NEP as the dependent variables. The significant level of the explanatory variables is at < 0.05 TOC, TN and chlorophyll were subsequently higher.
In contrast, at the lower region, the cryoconite grain size were coarser with a lower TOC, TN and chlorophyll content (Table 3). This observation corresponds with studies on Svalbard glaciers, where Kastovská et al. (2005) and Stibal et al. (2006) observed a high abundance of autotrophs and heterotrophs associated with cryoconite with a high proportion of fine sediments and with greater water retention time. The observed high organic content of finer cryoconite might reflect the greater nutrient availability for microbial activity than that of the coarser cryoconite at the lower site. Chlorophyll-a and TOC contents also increased with elevation, showing higher NEP values in the upper region (Fig. 5). The abundance of phototrophs and their potential productivity on glacier surfaces are often influenced by altitudinal gradients in mountain glaciers (Takeuchi et al. 2006;Takeuchi 2013), including the Himalaya (Yoshimura et al. 1997) where a general decrease in algal biomass is observed with altitude. These latter observations, however, were made on snow algal communities. Conversely, cryoconite holes present on the glacier's surface may provide a protective environment for these phototrophs to survive in the extreme conditions, depending on the local ice topography, shading, availability of irradiance ). This may explain why higher biomass and productivity were sustained on the Chhota Shigri at higher elevations, as discussed in detailed below. An important factor contributing to the enhanced organic matter accumulation in the upper regions of our experimental sites is likely the relatively low hydrological connectivity and gentle glacier surface slopes which reduce disruption of the cryoconite holes compared with those located in the middle and lower glacier regions studied here. Greater hydrological stability may lead to the formation of stable aggregates ) which support higher rates of biological activity, thereby enhancing the accumulation of autochthonous organic matter (Bagshaw, et al. 2013). From the statistical analysis, elevation (also related to the stability of cryoconite holes in the present study; "Physical and Chemical Properties of Cryoconite Holes") accounted for 16.3% while TOC accounted for 24.6% of the total variance in the Photosynthesis (Table 5). This observation suggests that the position of the cryoconite holes on the glacier surfaces has an influence on the primary productivity of cryoconite holes, and the consequent accumulation of organic matter. This is in alignment with a study by Stibal et al. (2012b), where enhanced microbial activity upslope on a glacier in Greenland resulted in increased organic matter accumulation because of less disturbance by flowing water on gentler ice surface slopes.
Another potential factor controlling the abundance of biomass production is the rate of Photosynthesis, which depends in part on the availability of photosynthetically active radiation (PAR) (Säwström et al. 2002;Cook et al. 2010). Although PAR was not measured in this study, the higher rate of Photosynthesis observed in the upper region in our experiments might be consistent with higher light availability for the autotrophs at a higher elevation compared to the lower elevations due to lower topographical shading. Multivariate statistical analysis also supported chlorophyll-a as being the most significant factor controlling the ecosystem production accounting for 39.6% (df = 1, F value = 15.7, p = 0.009) of the total variance in Photosynthesis (Table 5). The higher chlorophyll-a contents at the upper study sites are consistent with higher rates of Photosynthesis in these regions. The lack of a significant relationship between Photosynthesis and sediment thickness may also reflect light limitation in deeper layers of sediment (Telling et al. 2012).
The NEP rate of thin cryoconite sediment layers < 0.3 cm were also observed to be consistent with net autotrophic conditions, while at sediment thicknesses of ≥ 0.5 cm, cryoconite generally exhibited net heterotrophy, due to greater rates of Respiration (Fig. 4). This effect was pronounced at lower and middle glacier sites compared to the cryoconite in the upper region. RDA analysis also showed that TOC and sediment thicknesses of the cryoconite layer exert a significant control on NEP, accounting 35.8% (df = 1, F value = 9.7, p = 0.021) and 22.1% (df = 1, F value = 6.0, p = 0.04) respectively of the total NEP variation. It is also possible that the existence of various cryophilic invertebrates, such as copepods, tardigrades and rotifers reported in the Himalayan cryoconite holes (Kohshima et al. 2002;Zawierucha et al. 2021) may also contribute to the respiration although their contribution is not quantified here. The development of potential anaerobic microbial niche in thick cryoconite layers (~ 4-6 mm thickness) may also occur (Poniecka et al. 2018;Buda et al. 2021). From the RDA analysis, we found that sediment thickness explained 37.1% of the total variance in Respiration (df = 1, F value = 53.7, p = 0.001), and chlorophyll content explained 24.8% of the variance (df = 1, F value = 36.0, p = 0.002) ( Table 5). We note these observations were drawn from a small sample size, and therefore, need to be further tested with different sediment thicknesses along the altitudinal gradients with a larger sample size. However, a similar association between thin sediment layers, fine grain size and enhanced NEP and organic matter in cryoconite holes were also observed in the upper ablation area of Greenland Ice Sheet (Cook et al. 2010;Stibal et al. 2010), as well as in Svalbard glaciers (Telling et al. 2012) and suggests that thicker sediment layers may directly influence heterotrophic activity in Himalayan cryoconite. We note as well that cryoconite net ecosystem production is also likely be altered by seasonality. For example, a study on an Alpine glacier demonstrated that cryoconite holes are generally dominated by heterotrophs later in the season, while in the early melt period autotrophs like cyanobacteria dominated (Pittino et al. 2018). NEP on Chhota Shigri was lowest at the middle parts of the ablation zones with net heterotrophy predominating at the lower sites, which were accompanied by decreased chlorophyll-a concentration, organic matter, and coarser cryoconite particles. The decrease in chlorophyll-a concentration and organic matter down-glacier can be attributed to the formation of new channels which increased meltwater flushing on steeper slopes during the melt season (Takeuchi 2013) resulting in the break up of cryoconite holes. This rapid hydrological flushing of meltwater channels can result in a trend towards a net heterotrophic conditions (Hodson et al. 2007;Telling et al. 2012). We compare our measured rates of NEP with those measured using similar methods in polar glaciers and ice sheets in Table 6. These show that our measured NEP rate fall within the range of those observed in the polar glaciers and ice sheets, and the measured net Respiration and Photosynthesis rates in the present study are more similar to those measured from the Antarctic Ice Sheet. The low net Respiration and Photosynthesis in the present Himalayan cryoconite holes, compared to that of Arctic glaciers, could reflect their shorter lifetime on Himalayan glacier surfaces  which is explained in the following section. Table 6 Comparison of rates of net ecosystem production (NEP), photosynthesis (P) and respiration (R) with those of previous measurements of cryoconite in Polar Regions using similar method as the present study * NEP value was calculated from the mean differences of P and R reported in the respective study Location

Sources of organic matter in the western Himalaya
Since most research into glacier surface biological productivity is conducted in the Polar Regions, studies in the Himalayan region are very limited. Telling et al. (2012) from their earlier studies on Arctic glaciers proposed three possible sources for enhanced organic carbon content in glacier cryoconite. First, it may arise from long-distance deposition of rich organic carbon from outside the glacier to the cryoconite holes. Second, it may reflect autochthonously produced organic carbon in supraglacial habitats and re-deposited into cryoconite holes. Third, it may be due to very high NEP rates within cryoconite holes under favourable environmental settings such as thin sediment layers (Telling et al. 2012) or lateral thermal expansion of cryoconite holes, following maximum exposure of the cryoconite to solar radiation (Cook et al. 2010) which may result in greater organic matter production within the holes. In order to evaluate the potential organic matter accumulation mechanisms, here we estimate using our experimental data, the number of melt seasons it would take for organic matter and phototrophic biomass to accumulate in cryoconite holes at our two Himalayan glacier sites. The total number of melt seasons required to generate the observed TOC content (µg C per g − 1 sediment) of our cryoconite samples was assessed in a similar manner to Telling et al. (2012), as follows: Here, total carbon (TC) of the cryoconite (µg C g − 1 dry sediment) was taken to be approximately equal to the TOC content. Only samples with net autotrophic NEP values (µg C g − 1 day − 1 ) were considered for the calculation. The melt time is the typical duration of the melt season, which we assumed as 124 ± 13 days (mid May-mid September) for Western Himalayan glaciers as estimated by Panday et al. (2011). The number of melt seasons required to accumulate phototroph biomass was also calculated in a similar manner using Eq. (2), by substituting TC content for phototrophic biomass. Phototrophic biomass was estimated using a ratio of 1:47 between the chlorophylla concentration of cryoconite and the TOC content based on the estimation in phytoplankton species by Riemann et al. (1989). However, this estimation may not take into account all the extracellular polymers produced by autotrophs, as cryoconite may have a much higher ratio of organic carbon to chlorophylla . Caution must be exercised in interpreting these data, since the experiment performed was a snapshot study (done just for 24 h) in the ablation period only, and could not account for the rate of migration of the cryoconite particles, oxygenation and pH profiles within the cryoconite layer.
We estimate that the average time to form a thin layer of autotrophic biomass (approx. < 0.3 cm) on the glacier surface is 1.7 ± 1.5 years (with a range of 4 months − 3 years, with an outlier of 5 years) (Fig. 7). Within the same cryoconite hole, the average time required for organic carbon accumulation to account for the measured carbon contents is 68 years (range 11-74 years, with an outlier of 238 years) (Fig. 7). The above calculations show that during the in situ experimental periods, the overall net autochthonously produced organic carbon inside the cryoconite holes could not explain most of the accumulated organic content. The age of hydrologically connected cryoconite holes on the Chhota Shigri and Patsio glaciers is unknown. In Svalbard valley glaciers, it is estimated that the typical age of cryoconite is between > 1 year and several years, (Stibal et al. 2008). Hydrologically isolated cryoconite holes like those on the Greenland Ice Sheet, however, may have a lifetime of > 100 years and will be in a position to produce and accumulate organic matter better via autochthonous processes (Hodson et al. 2010a;Stibal et al. 2012a). In contrast much older carbon (6-10 ky) is found to accumulate over time in ice-lidded cryoconite holes on the Antarctic Ice Sheet which might be isolated from the atmosphere for much longer periods (Lutz et al. 2019). Compared to polar regions, cryoconite holes in the Himalaya are shallower and have a shorter life span due to high melt rates , therefore more frequent redistribution of the cryoconite across ice surfaces might be expected. This may explain the lower rates of Respiration and Photosynthesis observed in our studied cryoconite holes compared to Arctic glaciers and the Greenland Ice Sheets (Table 6). Since the cryoconite holes in the Himalayas are formed quickly and generally do not last long, most of the carbon inside the cryoconite hole most likely either reflects allochthonous deposition, or biologically produced organic carbon formed elsewhere on the glacier surface, for example, in surface ice via glacier algae Telling et al. 2012;Stibal et al. 2012a). We examine the potential source of the organic carbon and phototroph biomass below.
In the Himalaya, long-distance transportation and deposition of allochthonous organic carbon have been observed on snow and ice surfaces (Xu et al. 2013;Niu et al. 2017;Zhang et al. 2018). The Western Himalayan region, because of its remote position, is less influenced by long-distance transportation, therefore most supraglacial carbon content has been inferred to originate from in situ biological processes by photoautotrophic and heterotrophic activity, and deposition of carbon, derived locally from biomass burning of C3 vegetation (Nizam et al. 2020). Another source of organic matter in the Himalayan supraglacial environment is that from snow and ice algal communities (Yoshimura et al. 1997. Various diverse microbial communities, such as cyanobacteria, fungi, and plankton, thrive in snow, moraine lakes and stream water of Himalaya (Liu et al. 2011;Kammerlander et al. 2015;Sanyal et al. 2018;Azzoni et al. 2018) could be potential organic carbon sources in the region. An abundance of phototrophic cyanobacteria such as microcoleus and oscillatoriales has been reported in cryoconite from Asian glaciers during the later melt season (Segawa et al. 2017). On the Greenland Ice Sheet, snow and ice algae constitute a majority of the supraglacial microbial community and dominate the net primary production after the onset of melt season (Yallop et al. 2012;Lutz et al. 2014;Stibal et al. 2017;Williamson et al. 2018Williamson et al. , 2019Williamson et al. , 2021. This algal biomass retained on glacier ice surfaces has the potential to sustain cryoconite holes by providing organic carbon and nutrients to other microbial communities Wadham et al. 2010;Antony et al. 2014;Lutz et al. 2015;Williamson et al. 2018). Additionally, secondary heterotrophs surviving on glacier surfaces may further modify the organic carbon deposited thereby producing complex organic carbon through in situ processes (Sanyal et al. 2018;Nizam et al. 2020). Considering the estimated time taken to produce organic carbon (average ~ 68 years) on our study glaciers, and the inferred short life of cryoconite holes in the Himalaya , we infer that a significant proportion of the organic carbon in cryoconite might originate outside of cryoconite holes through surface ice microbial production redistributed in the supraglacial environment and incorporated with the freshly formed cryoconite holes or from vegetation biomass redistributed by wind (Nizam et al. 2020). This interpretation would reconcile our lengthy times for cryoconite TOC to accumulate and support the recent finding from biomarker and geochemical tracers (δ 13 C, Pb isotopes) that TOC in cryoconite holes in the Western Himalaya has a strong biological origin (Nizam et al. 2020).

Future implications
Potential microbial activity on the surfaces of two largely debris-free Himalayan glaciers has been revealed for the first time through in situ NEP and Respiration measurements. Our study provides a significant first step in evaluating the potential influence of biological activity on the physical behaviour of the glaciers in Himalaya. Since Himalayan glaciers share a similar climatic environment to that of the Polar Regions, biological activity observed on Himalayan glacier surfaces might also play a significant role in driving the glacier melting by absorbing summer solar energy. The organic matter accumulated on the glacier surface, through primary production, may lead to a positive feedback on glacier melting (the so called "bio-albedo" effect) Lutz et al. 2014;Cook et al. 2017;Hotaling et al. 2021). Although glacier mass loss due to the bio-albedo effect is not well quantified, a study by Williamson et al. (2020) reported that glacier algae may be responsible for up to 75% of the variability in albedo in the Southwestern Greenland Ice Sheet. While ice algae are important in surface albedo reduction on an ice sheet scale (Yallop et al. 2012;Lutz et al. 2014;Stibal et al. 2017;Williamson et al. 2018;Tedstone et al. 2019), the importance of ice algae on valley glaciers like Himalayan glaciers are unclear. It is, therefore, important to focus on the microbial communities both on the ice surfaces as well as cryoconite holes further in future. Apart from the potential physical changes on glacial surfaces due to microbial activity, the organic matter deposited on the glacier surface within cryoconite holes and dispersed on the glacier surface may result in the export of dissolved organic matter and nutrients to downstream ecosystems, thereby changing the activity and function of the recipient microbial community. Therefore, in light of the important ecosystem processes carried out in this extreme environment, future research on Himalayan glaciers should incorporate the potential role of microbial activity on the glacier mass loss and its possible implications for downstream ecosystem.

Conclusion
Potential microbial activity in the cryoconite holes of two largely debris-free Western Himalayan glacier surfaces, Chhota Shigri and Patsio glaciers, was measured for the first time. Diverse microbial and environmental variables including organic matter content, chlorophyll-a, altitude, grain size and stability of the cryoconite were studied to identify and understand the factors controlling the NEP. A significant association between organic matter (TOC % and TN %) with chlorophyll-a content (r = 0.94 and 0.95 respectively = 10, p < 0.05) and a general increase in the fine mineral fraction of cryoconite from lower and middle to the upper sites in Chhota Shigri glacier was observed. Net positive autotrophy dominated the cryoconite holes in the upper ablation region at Chhota Shigri and at Patsio glacier, which hosts relatively stable cryoconite holes with a relatively high chlorophyll-a content. The latter enables microbial communities to carry out photosynthesis in the supraglacial environment of the Himalaya. Net heterotrophy associated with coarser grain sizes, reduced organic matter and chlorophyll-a content was observed in the lower ablation zone, which potentially reflects the disturbance of cryoconite holes by frequent flushing by meltwater channels. Multivariate statistical analysis using RDA indicated that sediment thicknesses were a significant factor controlling the rate of Respiration (37.1%; df = 1, F value = 53.7, p = 0.001) in cryoconite holes in Chhota Shigri. Chlorophyll-a content of the cryoconite holes significantly controlled the rate of Photosynthesis (39.6%; df = 1, F value = 15.7, p = 0.009). While the rate of NEP were significantly controlled by the TOC content and sediment thickness of cryoconite holes (35.8% ; df = 1, F value = 9.7, p = 0.021 and 22.1%; df = 1, F value = 6.0, p = 0.04) of the total variance in the experiments performed in a limited time. Phototrophic biomass in the cryoconite holes (sediment thickness < 3 cm) was estimated to take 1.7 ± 1.5 years to accumulate, supporting its production via autochthonous processes on the glacier surface. However, the organic carbon in the same cryoconite holes took an estimated 11 to 70 years to form, indicating that allochthonous and/or autochthonous organic carbon formed elsewhere on the glacier surface, such as in surface ice, may contribute to the accumulation of organic carbon in cryoconite holes.